Coolant issue - radiator empties itself to the overflow

[crazydriver81]

Well, where to start... I have an issue with the coolant and need some advice/experience of the bunch of people here.


The issue: when driving for longer distances the overflow bottle get's filled, I loose about 200ml of coolant and there are bubbles coming up in the oveflow bottle. This only happens, when engine was completely cooled down before driving.

When I have reached my destination, I usually refill the radiator after letting the car sit for an hour. The radiator is full then and the coolant bottle is on min marking.

When the car cools down completely (e.g. sitting over the night) the overflow bottle empties itself towards the radiator.

I have found that this happens more, if I use to drive with under boost a lot. The effect is nearly not there, when I drive with 130km/h constantly on the motorway. In any way, oil temps are moving between 90 - 105 °C, but not above.


The car:
- alloy radiator from the group buy with 1.3 bar rad cap (was there with the std. radiator as well)
- HKT low temp thermostat 71°C
- new coolant (green, no OAT stuff)
- no visible leaks
- ECU temp sender changed
- radiator fan changed
- changed the radiator cap - no improvement
- full cambelt service with brand new water pump 3 years/5000kms ago

All parts have been on the car for max 5000kms and I had this issue for a longer time now. We logged the temperatures with EvoScan and the car reaches 100°C on idle easily, fans kick in as expected. I did not had the time yet to check temperatures while driving but even with 20°C ambient temperature this happens.

We did perform a check, whether there are residues of exhaust gases / water and did not detect any residues of water in the exhaust gases yet, nor did we found exhaust gases in the cooling system. There is no white smoke, nor do I have any residues of water in the oil.

Next is to perform a coolant system test if there are any (visible) signs of leaks.


On a sidenote: I found some oil build up, running down the intercooler pipe (after Y-pipe) after 400kms driving - although I cleaned it before.


Has anybody heard of hairline cracks on a 6A13TT?
Can a faulty turbo cause this to happen? (e.g. pressurize the cooling channels)
Any other ideas, as the head gaskets seem to be fine?

[evo1rs]

I thought that was the normal operation for a sealed cooling system with expansion tank. When hot, the coolant expands filling the tank, when it's cooled it contracts, drawing the coolant back into the rad. If you're constantly loosing coolant though then it must be going somewhere

[Davezj]

This sound like the problem I was having, and think that mine is fixed. And it was a faulty rad cap that caused the issue. I went through 5 rad caps before I found one that worked properly.
I suspected a faulty rad cap, so I bought a dirct replacement from eBay this did not fix the issue, so 2 x 0.9 bar and 2 x 1.3bar rad caps later and my issue seems to be fixed. With the last 1.3 bar rad cap. But I only think it is fixed, as when I really really boot it, I get some coolant leaking out of the expansion bottle, which settle back down to just above min in the expansion bottle when it cools down so not too bad but I have my niggling thought that there may still be an issue there.

The rad caps bought from eBay were china jobs but had logos like blitz, hks and ralliart, all of hem just let water flow out to the expansion tank and held no pressure at all. Completely useless. The last one I bought was unbranded, no stikers and works.
So don't buy the cheap crap from china, unless you want to by 2, 3, 4 times before you get a good one. It is just not worth it.

To check what is happening you could try and fit a pressure gauge to the cooling system to see if it goes really high. Or you could do a pressure test on the cooling system with pressure test cap and pump air into the system and check for leaks.

[wintertidenz]

Also make sure you get all bubbles out of the system when bleeding it - I had this issue till I re-bled my system and found an air bubble.

[Kenneth]

First replace the radiator cap. I would see if you can borrow one from someone, I love to do my problem solving free

If you use a good radiator cap and still get the issue, it could be a head gasket failure. In fact, if it ends up not being the radiator cap, I would be say it is classic signs of a head gasket failure.

[Davezj]

Also make sure you get all bubbles out of the system when bleeding it - I had this issue till I re-bled my system and found an air bubble.

yes bleeding the air out fully is a great benefit.
i found a funnel with the same diameter nozzel as the the radiator cap works very well. just run the car until the coolant is just below boiling point, let it cool down and all the air will be pushed out. i found this to be a more reliable way than using the bleed screw as air can et trapped in the system and you need to rev the car to push it round to the rad cap. the funnel gives the expansion space and the reservoir to allow the air to come out and be replaced with more coolant automatically.

[adaxo]

@mcosworth visit me yesterday and he describe very similar issue to yours @Crazydriver81 and @Davezj I say to him to start sorting this out by changing rad cap, what is odd IMO that you all running aftermarket rads, Stefan & Maciej golpher and Dave some evo one, so my bets will be on dodgy fitting/hole/seal/whatever has contact with rad cap in aftermarket rads.

[Davezj]

this is true adam, but my issue start when i had standard rad fitted, and i overheated the cooling system very very badly. So i am not surprised i had issues later on. the alloy evo rad was one of the things i tried to combat the problem.
but that does not rule out the possibility that there is now a sealing issue.
i would like to point out there are 2 different rad cap options when you buy replacement caps. there is a standard 9mm depth spring and a 15mm depth spring. my evo rad requires a 15mm deep spring and the factory rad requires a 9mm deep spring.
so make sure you buy the correct cap.
i only found this out due to one of the ebay adds i looked at offered both option and you had to specify which you wanted at the checkout stage.

i will have to take a picture of my funnel in place when bleeding my rad. just so you know what actually mean.
is maciej back from Poland now. if so i will have to bob round a drop your crank off and then go round to his and drop the top and bottom rad hoses off.

[Ghost_2008]

This sounds exactly the issue I was having for a very long time. The engine run spot on the whole time, no issues noise etc.... If I drove short distance 5-10 miles no issues the cooling system functioned as it should, however long journeys would cause overheating but only once I had stopped.

When inspecting the overflow bottle I would find the bottle full and bubbles from the overflow pipe... In an effort to resolve the issue I replaced the cap, water pump, thermostat, all pipe work and censors...

The only possible cause in the end, gasket. I believe the head gasket failed on the exhaust side allowing exhaust gas the enter the cooling system. The gas expands and over a long journey would empty the rad enough to cause overheating..... I have recently replaced the engine, for a full built unit, I'm still running it in, but I just got home from a 300 mile round trip with no problems.....

I hope this isn't the same for you but your initial description is almost identical...... I would add however that I also replaced both turbo'ş with the engine, both fully rebuilt, look brand new in fact.... Also I didn't actually check for gases in the coolant......

[crazydriver81]

I just can emphasize that I had the issue for a longer time now, even before the alloy radiator went in. In fact, this was the reason for me to investigate radiator options and decide to buy an alloy rad. I doubt, thas this effect is down to a dogy (aftermarket) alloy radiator. I also doubt that this has something to do with the radiator cap - as it was supplied along with the alloy rad and (hopefully) the manufacturer has supplied a sufficient quality.

I have tested with a different radiator cap - and still had the same problems. I also feel that the effect has gone worse now - but this can also be due to the fact that there is simply more coolant in the whole system, due to the double core of the alloy rad.


I have an appointment in the garage, to perform a leak down test. If it's really the head gasket, there must be signs of exhaust gases, which are easily traced back with the leak down test. I will report back then. However, I have no signs of oil in the water, nor signs of water in the oil. I hope it will not be the head gasket. Otherwise this will be a final decision for the car.

Further I bought a D1 Spec radiator cap with built in temp gauge - do get an idea of the temperatures of the coolant in the radiator.

I will re-bleed the cooling system and also need to build a kind of flow check for the upper radiator hose - to determine if there is enough flow of coolant. Though I don't have an idea yet how to do that...

But an overheating engine - wouldn't that be seen on oil temps too? As said, never above 105°C...

[crazydriver81]

I also understand that bleeding the system is essential, due to the small water pipes to turbos and etc - as air can get trapped easily.

However, after my drive at the weekend, the system should be bleeded, as it took back a lot of coolant from the reservoir tank while cooling down over night. Air should be out then I guess.

@mcosworth - have you installed the radiator yourself, as I did?

[mcosworth]

Yes, I fitted it myself.

[crazydriver81]

How is the cooling system expected to work? What I mean by this is: engine warms up until thermostat opens. But how is the flow of the coolant then? Engine - Top hose - radiator - lower hose - engine? Or does thermostat open and the flow towards radiator is via the lower hose?


I definitely have an air lock as you can hear it when pressing the cold top radiator hose. Also my top radiator hose is warm and tight, but i can squeeze the lower hose and it's just a bit warm? After a Short Drive it's still the same - there is a temp difference between top an lower rad hose.

[GalantOnly]

Had the same issue with this guy:

http://www.clubvr4.com/forum/showthr...-leaking-water

Tried EVERYTHING. Negative. The last step was the head gaskets. When I took the heads off, I couldn't see anything wrong. But two fresh Camskill items solved the problem. Sorry, Stefan...

[crazydriver81]

@GalantOnly - I can remember this case. Though there was no final conclusion from the user.

Next steps are: leak down test for exhaust gases in coolant / pressure test / removing thermostat temporarily

And removing the air which is in the system!

If it is the Head gasket, i will not fix that.

[crazydriver81]

Anyone who knows how the flow for the coolant is?

[Davezj]

For completeness here we go.

With he thermostat closed it just goes round the engine.
When thermostat opens fully it block off the flow round the engine and direct all the water to the top rad hose, which then flow down through the rad and out the bottom hose back to the engine.

So I would expect the bottom hose to be cooler than the top hose as the rad is taking the heat out of the water. The hard ness of the top pipe verses the bottom pipe is a bit odd, a you say the water pressure in the entire system should be the same as it is a loosed system, but if the water is cooler in he ottom hose then the pressure should be less, plus the water pump will be sucking the water from this area and forcing it towards the top hose.

You could set up a flow test by using a glass tube inline with the top rad hose and you can see the flow in the glass tube.you buy these or get yourself a large test tube and cut it down to a tube.

[crazydriver81]

For completeness here we go.

With he thermostat closed it just goes round the engine.
When thermostat opens fully it block off the flow round the engine and direct all the water to the top rad hose, which then flow down through the rad and out the bottom hose back to the engine.

So I would expect the bottom hose to be cooler than the top hose as the rad is taking the heat out of the water. The hard ness of the top pipe verses the bottom pipe is a bit odd, a you say the water pressure in the entire system should be the same as it is a loosed system, but if the water is cooler in he ottom hose then the pressure should be less, plus the water pump will be sucking the water from this area and forcing it towards the top hose.

You could set up a flow test by using a glass tube inline with the top rad hose and you can see the flow in the glass tube.you buy these or get yourself a large test tube and cut it down to a tube.

Thanks Dave. I was just not sure if the water flows out via the top or bottom hose. So the return is the lower hose.

[crazydriver81]

I am back from the garage now. I was bleeding the cooling system - which is a true PITA job.

The cooling system had a lot of air trapped in the lower radiator hose. You could also hear water/air when pressing the cold radiator hoses. Finally, I added 600ml of coolant, before I could get all air out - which in return means I had around 10% air in the cooling system!

I took the car out for a spin and first signs are good. I also gave the car some beans once warmed up and the coolant level in the overflow bottle stayed as it was before - so no coolant was pushed to the overflow bottle. Also the temperatures of the radiator hoses were nearly identical now.

I will have a spirited run in the car at the weekend, with motorway driving, to see how it goes. Either the solution was that easy or problem it will return.

[crazydriver81]

@mcosworth: Please check your rad hoses when engine is cold. Can you hear water moving around and can you easily press the hoses? It's likely that you will have air in the system.


How I got the air out:
- open rad cap with engine cold
- fill the overflow bottle to max and close it
- take your bare hand and "close" the radiator filler neck
- press the upper radiator hose and you should hear bubbles in the overflow bottle
- if you depress the radiator hose now, it takes back in coolant from the overflow bottle as you have created a vacuum
- repeat for upper/lower hose until no bubbles are left (upper hose - lower hose -upper - lower - ...)

It took me half an hour to do this and as a result the radiator was filled to the max.

The design of the Golpher radiator is different to the stock radiator design. I suspect that the air cannot get to the highest point and escape as easy - so proper bleeding seems to be difficult and crucial.